What happens at 2000 RPM?

[ijonsson]

Hi,

Just arrived home from annual meeting of TR club Sweden, really nice event, and I made a total 800 km of driving over the weekend (links to some pics from the event below). To my surprise I made 29 mpg on averring 65 mph on a 125 miles stretch on a 2+1 road. AFR read approx 14,2 most of the time, short lean peaks to 17-18 on accelerating and almost no pinking.

BUT, on 3, 3+OD, 4 and 4+OD trying to accelerate through 2000 RPM engine misfires quite bad at exactly 2000 RPM, noting below, noting above. Is it distributor problem, or might I've involuntarily caused this by fiddling with the MU set screws trying to make the engine run as good as it can get. No big changes , max a 1/4 to a 1/2 turn on the set screws.

44 TR 2, 3, 4, 5, 6, 7 and 8 gathering in front of a small castle

On the road

86 year old lady driving her TR2 (son at her side)

Edited August 4, 2014 by ijonsson

[snowric]

Hello,

 

probably no help whatsoever from me but if it is at 2000 rpm as you will experience 2000 rpm at different inlet vacuums isn't it more likely a dizzy problem as no vacuum input on dizzy!

 

I wondered if I could ask you to send me a short mail on how you adjust the MU and measure AFR? I have read the various tech docs so am aware of how the slopes vary with changing the differetn nuts and how the cam changes the slope.

 

But i don't know whether turning the top set screws clockwise results in leaning or richening and what is the easiest way to insert a port into the manifold, i use a standard.

 

My car (a TR5 with uncertain or tired cam) runs >7mmHg (8-9mm) at idle and is overfuelling.

 

It is also doing about 15mpg, albeit with lots of town driving.

 

I am a happy tinkerer but need a bit of help to understand before I take the black top off !

[ijonsson]

Hello,

 

probably no help whatsoever from me but if it is at 2000 rpm as you will experience 2000 rpm at different inlet vacuums isn't it more likely a dizzy problem as no vacuum input on dizzy!

 

I wondered if I could ask you to send me a short mail on how you adjust the MU and measure AFR? I have read the various tech docs so am aware of how the slopes vary with changing the differetn nuts and how the cam changes the slope.

 

But i don't know whether turning the top set screws clockwise results in leaning or richening and what is the easiest way to insert a port into the manifold, i use a standard.

 

My car (a TR5 with uncertain or tired cam) runs >7mmHg (8-9mm) at idle and is overfuelling.

 

It is also doing about 15mpg, albeit with lots of town driving.

 

I am a happy tinkerer but need a bit of help to understand before I take the black top off !

You have a PM

/Inge

[Peter Cobbold]

Inge,

The AFR sensor certainly removes uncertainty about fuelling. But if the engine misfires at all the AFR reading in not valid anymore. So we dont know what the AFR is during the 2000rpm misfire. Did the AFR show a trend towards leaning just before the misfiring?, and a richening recovery after?

Or vice versa - over-rich misfires can occur.

 

2000rpm is about halfway up the centrifugal advance curve so I would check the spark timing with a strobe. That can be done with the car stationary.

 

Perhaps you dont detect the 2000 misfire in 1st and 2nd gears becasue the car accelerates faster in low gears through 2000rpm.(wind resistance and tyre drag are less than at speed) So drive up a steep hill and the 2000 misfire may appear in 2nd gear too.

 

Peter

[ijonsson]

Inge,

The AFR sensor certainly removes uncertainty about fuelling. But if the engine misfires at all the AFR reading in not valid anymore. So we dont know what the AFR is during the 2000rpm misfire. Did the AFR show a trend towards leaning just before the misfiring?, and a richening recovery after?

Or vice versa - over-rich misfires can occur.

 

2000rpm is about halfway up the centrifugal advance curve so I would check the spark timing with a strobe. That can be done with the car stationary.

 

Perhaps you dont detect the 2000 misfire in 1st and 2nd gears becasue the car accelerates faster in low gears through 2000rpm.(wind resistance and tyre drag are less than at speed) So drive up a steep hill and the 2000 misfire may appear in 2nd gear too.

 

Peter

I'm sure you are right about first and second gear. I'll also check timing again, I remarked in another post that the engine was running OK with 18 deg advance at tickover, but I've retarded the timing a bit since. I'm unfortunately not able to make the AFR monitor store data for reasons unknown but I'll keep an eye on the monitor next time I take the car for a spin. I still think it's a bit strange that misfire is in such a narrow range, it's 2000 +/- 50 rpm:s, always the same.

[Peter Cobbold]

I still think it's a bit strange that misfire is in such a narrow range, it's 2000 +/- 50 rpm:s, always the same.

You mean it misfires only between 2000 and 2050 rpm? Sounds like it might be a resonance somewhere.

Is that only a wide open throttle? or at intermediate throttles? and if stationary?

Peter

[ijonsson]

You mean it misfires only between 2000 and 2050 rpm? Sounds like it might be a resonance somewhere.

Is that only a wide open throttle? or at intermediate throttles? and if stationary?

Peter

Yes, the problem is only between 1950 and up to 2050 rpm, At wide open it almost dies in that interval, misfires at intermediate throttle and runs OK at constant speed. I'm avoiding this interval by flicking OD in and out till problem solved.

I also can monitor fuel pressure and it's rock solid 104 psi at all loads (measured at the PRV, PRV comes from Revington by the way), the AFR system I'm using is SLC PP2 (from 14point7) but unfortunately something is wrong with the units data storage so I can't download any trends.

Edited August 4, 2014 by ijonsson

[Peter Cobbold]

Yes, the problem is only between 1950 and up to 2050 rpm, At wide open it almost dies in that interval, misfires at intermediate throttle and runs OK at constant speed. I'm avoiding this interval by flicking OD in and out till problem solved.

I also can monitor fuel pressure and it's rock solid 104 psi at all loads (measured at the PRV, PRV comes from Revington by the way), the AFR system I'm using is SLC PP2 (from 14point7) but unfortunately something is wrong with the units data storage so I can't download any trends.

That is really weird. I cannot see the PI causing that. Nor would timing errors be that 'fine tuned' with rpm.

Could it be a faulty rotor arm loosing contact with the cap contacts, but only when the engine is being accelerated?

Peter

Edited August 4, 2014 by Peter Cobbold

[snowric]

Peter,

 

Could the poster loosen the clamp and alter the distributor timing as someone accelerates the throttle through 2000 rpm to see if it is a timing problem ?

 

Snowy

[ijonsson]

I'll check the distributor and change the timing ( I have a strobe light) I also have a spare rotor and cap to test and see if it makes any difference, but as Peter said, "really weird" I've never experienced a fault like this before. If I can't cure it, I'll probably send the distributor to be refurbished when driving season is over here in Sweden, to at least eliminate distributor as a cause.

[Guest ntc]

Check the earth wire on the baseplate .

[Peter Cobbold]

Inge,

Yes as Snowy suggests. And measure the timing and set it to 12BTDC static ( the normal 9.5 compression setting).

More questions:

Did the fault develop suddenly ( something breaking) or gradually ( something wearing or loosening).

Points gap OK?

Is the disy firmly clamped down against its end-float shim?

Points plate correctly held by the spring from the capsule? http://www.tr-register.co.uk/forums/index.php?/topic/35961-why-was-the-distributor-turned-so-much/

Do the centrifugal advance springs feel right when you twist the rotor arm?

What does the misfire sound like? banging in exhaust? a 'hollow' exhaust note? inlet backfires?

Peter

 

Edit: Check Neil's earth wire first. That could have a 'resonant' behaviour, I think.

Edited August 5, 2014 by Peter Cobbold

[RogerH]

Hi Inge,

the 2000rpm is too fine for normal adjustable items (timing, fuel, air etc) in my opinion.

However it could easily be a resonance that is disturbing some wiring to the coil/dizzy etc.

 

Check ALL the engine wires for security.

 

Inside the dizzy is the LT wire that take the coil LT volts to the contact breaker.

Remove this wire and inspect the two ends where they are soldered to the crimped tags.

On my 4A the crimping was on the insulator and the solder had broken away from the tag - but barely visible.

 

Roger

[Mike G]

Hi
I had an identical problem, with a flat spot around 2000 RPM. I tried everything, but could not pin down the problem. I eventually took the car to a Lucas PI specialist. It turned out he had come across this problem on a TR5. He had the 5 on his rolling road, for 2 days but could not find the reason for the flat spot.
Eventually he started to change out the PI components one at a time. When he changed out the flexible fuel hose to the metering unit, the flat spot disappeared. He changed the fuel hose on my car and again the car revved smoothly through the 2000-RPM rev range. There appeared to be no problem at all with the hose, no kinks or blockage. He told me that the hose he removed from the TR5 looked fine as well.
If you cannot trace the problem with your car, give it a try.

Mike G

[ijonsson]

Yes, it might be a resonance problem as Peter suggested, I've been revving up and down in the garage (thunderstorm and rain outside doesn't invite to driving) and think I can hear an "out of place" sound just around 2000 rpm. I'll do some driving when weather is better after changing timing and replacing some bits in the distributor.

To answer some of the questions above, I can't find any badly connected wires, distributor is firmly clamped, advance springs works on rotor, point gap is OK, checked with timing light etc.

And the sound of it, no banging, pinking or backfire, it's more like no firing at all, engine is "stuttering".

It's problems like this that keeps you on your toes, right?

[RogerH]

Inge,

look at that dizzy LT wire.

 

Roger

[Peter Cobbold]

Inge,

Does the timing strobe fail when the engine stutters?

 

If the strobe is fine, then we should look at MikeG's post, #14.

Why did changing the rubber high pressure hose to the MU eliminate a flat spot? (on two TRs - not a fluke).

I suggest it is the compliance of the new hose that was different from the old. And the old hose allowed a

'water hammer' type of hydraulic resonance*.

On your car you have fitted a take-off to monitor fuel pressure.

Maybe that is acting to produce a resonant water hammer at 1950-2050rpm, worse when the fuel flow is greater.

So I suggest blanking off the psi-gauge take-off. Or better, refitting a standard hose.

 

I may well be wrong,but peculiar faults need peculiar answers!!

* http://www.engineersedge.com/fluid_flow/water_hammer.htm

 

 

Peter

Edited August 6, 2014 by Peter Cobbold

[ijonsson]

Hi,

Can't provoke the problem parked in the garage, it only occurs during acceleration on gear 3 and 4 so I can't see anything on the strobe. The take off for fuel pressure transmitter is at the PRV (no hoses involved) so I don't think that's the problem. Not long ago I put in a refurbished MU (Prestige) and of course loosened and tightened that hose and working on old rubber hoses might cause problems, so it could be worthwhile to order a new hose just for eliminating the hose as the cause.

I'll also take an extra look at the LT wire.

Thanks for all suggestions, the members of this forum are fantastic!

[Peter Cobbold]

Inge,

The hose would I think need to be stiffer than standard to support resonance. Decades old rubber could be harder. Or if it has been fitted slightly twisted. Did the fault appear immediately after disturbing the hose? Try refitting your present hose but ensuring it is completely free of any twist.

Peter

[ijonsson]

Finally I found out why I didn't get any real data from the AFR-logger, USA uses point as decimal separator and Sweden uses comma. Had to set up a laptop to use point and now I get trends, now I only have to figure out what they mean . Enclosed a graph I guess is from the worst moment of misfiring. I think I haven't used quite the right multiplier for rpm but what surprises me is that the rpm-curve makes dips to zero which would indicate loss of power to the coil, because that's where I'm picking up the signal. So, Roger, you suggested LT lead (in which I can't find any problem) as a possible cause. Could you please explain how you think the problem occurs, always at the same rpm.

Blue=RPM, Red=AFR, Purple=Acceleration/deacc,Green=fuel press, Blue=volts (to logger), Yellow=oil temp

[oldtuckunder]

Hi looking at the time line if that is seconds then the RPM trace shows no RPM for upto 30 secs, which would be more like a grinding halt than a misfire. I expect what is happening there is that some data loggers have a poor sampling method and the more inputs you give them the more exagerated it gets. I suspect what is happening is that in sampling the RPM signal its hitting a lot of misses and not many hits and calculating zero RPM. Given that the AFR is getting richer at these points and the acceleration line is going up, that would confirm that you were accelerating through the misfire, and the lean spike after the RPM comes back indicates misfire cleared and all cylinders firing.

Where is the AFR sensor mounted?

If your picking up the RPM signal from the negative coil terminal (which I think is correct) then I think what you are seeing is a failure of the signal going to ground not a loss of power to the coil. This would tend to back up Rogers suggestion.

Do you have a condensor in the distributor or an externally mounted one?

Possible condensor breaking down?

Alan

[RogerH]

Hi Inge,

my duff LT wire simply gave missing big time until it failed completely.

My connection with your problem is that it may be a resonance causing the make/break condition.

 

If yours is fine then that is one area to ignore.

 

Roger

[ijonsson]

Hi looking at the time line if that is seconds then the RPM trace shows no RPM for upto 30 secs, which would be more like a grinding halt than a misfire. I expect what is happening there is that some data loggers have a poor sampling method and the more inputs you give them the more exagerated it gets. I suspect what is happening is that in sampling the RPM signal its hitting a lot of misses and not many hits and calculating zero RPM. Given that the AFR is getting richer at these points and the acceleration line is going up, that would confirm that you were accelerating through the misfire, and the lean spike after the RPM comes back indicates misfire cleared and all cylinders firing.

Where is the AFR sensor mounted?

If your picking up the RPM signal from the negative coil terminal (which I think is correct) then I think what you are seeing is a failure of the signal going to ground not a loss of power to the coil. This would tend to back up Rogers suggestion.

Do you have a condensor in the distributor or an externally mounted one?

Possible condensor breaking down?

Alan

Hi,

the scale is sample points, I forgot to select seconds, so the "downtime" is between 2 to 3 seconds. The Lambda sensor is mounted just after where the downpipe goes horisontal and where I made an X connection of both pipes so I can measure all cylinders.http://www.ijonsson.se/triumph/photo_diary.php?day=motor&folder=motor&id=IMG_0579 . Condenser is new and of "best quality", I've also tried an old one without any improvement.

Here is another screen dump cleared up and x-axis in seconds

[Peter Cobbold]

I dont see a coil or sparking fault: if the engine stopped sparking completly the AFR trace would go sky high lean. It doesn't - even during the rpm trace outages. And a dead, sparkless, engine sounds and feels very different from a misfiring engine.

I reckon the rpm trace zeroings are artefacts.

 

 

Where did it misfire? - where does the full throttle fault at 1950-2050 rpm show up? I cant see it.

 

If some injectors are failing then the AFR reading will go lean. One injector dropping out completely will lean by one sixth. So an AFR 15 with all working, will go to AFR16.7. Two injectors out will take 15 to 20. There are excursions to 17 and 19 on the traces.**

But you describe an almost total loss of power at full throttle - at 1950-2050 rpm- which does not sound like a couple of injectors dropping out. If they all dropped out the AFR would go skyhigh.

 

Can we see a trace that shows the 1950-2050rpm fault at wide open throtltle when you feel the power drop. A single run from tickover to 3000rpm at wot in 3rd gear will be much easier to interpret.

 

I think I'd raise the fuel pressure to 115psi to see if those AFR lean swings clear up. 104 might not be 104, depends upon the sensor accuracy. Trace reads 101, lower than spec.

 

Peter

 

** If the throttle was at wot all the time, those excursions to 17 and 19 are not lean spikes caused by fuel depositing on manifold walls. But if the throttle was being opened and shut they could be lean spikes. A constant-wot trace is essential to be able to diagnose the AFR.

Edited August 6, 2014 by Peter Cobbold

[oldtuckunder]

IMHO this has got to be electrical look at the RPM trace at around 246 secs, there is no way that an engines revs can fluctuate like that, but as the RPM reading is coming from a spark induced signal, that fluctuation is exactly what you would see from a misfire and whilst about 200 rpm higher that reported is in the correct range.

 

Also the misfire would show up as rich not lean as unburnt fuel is getting through. NB some of that does depend on if a wideband or narrow band sensor is being used, which I don't think we know, but I'm guessing wideband.